2982
PETER G. BOWERSAND GERALD B. PORTER.
Fluorescence and Phosphorescence of Hexafluoroacetone Vapor
by Peter G. Bowers' and Gerald B. Porter Department of Chemistry, Unitersitv of British Columbia, Vancouaer, Canada
(Receized M a y 16, 1964)
The emission spectrum of hexafluoroacetone consists of both fluorescence and phosphorescence, of which the latter is quenched by trace quantities of either biacetyl or oxygen. From a quantitative study of the emission of biacetyl, direct and sensitized by hexafluoroacetone, and of the emission of hexafluoroacetone, the following quantum yields have been obtained at - 78' : phosphorescence, 0.51, and fluorescence, 0.05.
While there is good evidence from studies of the photodissociation of hexafluoroacetone (HFA) that the excited triplet state is involved,2a~b no phosphorescence was reported by Okabe and Steacie,3 who originally measured the eniission spectrum of this molecule. This contrasts with both acetone and trifluoroa c e t ~ n e , where, ~ ~ , ~ except at elevated temperatures, most of the eniission is phosphorescence. Besides the fact that the lifetime for phosphorescence is several orders of magnitude greater than that for fluorescence, the two kinds of eiiiission may be readily distinguished by observing the effect of oxygen. In these siniple ketones, where the upper state to which the niolecule is excited is I(nn*), it is well known that oxygen has little effect on the fluorescent emission from this state, but efficiently quenches phosphorescence froin the triplet state " n ~ * ) . It was felt that phosphorescence from HFA might best be recognized a t a very low tenlperature, where competing thernial dissociation of triplet state molecules would be coiiipletely inhibited. This paper reports the result of such a study.
Experimental Hexafluoroacetone gas, supplied by Allied Chemical, was outgassed for 2 hr. in a LeRoy-Ward still5a t - 12,5' and collected at - 110'. It was outgasscd for a further 2 hr. at -196' just prior to use. A sainple prepared from HFA hydrate (Merck of Canada) gave siniilar spectra. Eastiiian Kodak biacetyl was fractionally distilled on the vacuulii line and a luiddle fraction collected. Saniples for use were thoroughly outgassed. Oxygen was prepared by heating A.R. grade potassium permanganate. The T-shaped fluorescence cell for work at 25', Th,e Joiarnal of Physical Chemistry
4 cm. long and 2 cm. in diameter, was of Pyrex with three quartz windows attached with an epoxy resin. This cell was blackened on the outside and used without therniostating. The low temperature cell was similarly constructed; it was housed in a light nietal casing insulated with paraffin wax (1 cni.) and polyfoain plastic (2 cni.). During a run, the cell walls were completely surrounded by pulverized Dry Ice, which was frequently stirred to ensure good thermal contact. Evacuated guard tubes were attached to the cell windows to prevent condensation. HFA was dosed into this cell at room temperature, and 20 min. was allowed for cooling after adding Dry Ice. Light froni an Osraiii HBO 200 iiiedium pressure niercury lanip passed through a Bausch and Loinb grating monochromator and a Corning 9863 filter, and thence, as an approximately parallel beam, along the axis of the cell. The window on the T-arm of the cell faced the entrance slit of a Hilger f/4.4, D285 spectrometer with glass optics, Fluorescent light, after passing through the spectrometer, was measured with an RCA 72G5 photomultiplier. Emission spectra could be recorded between 7000 and 3800 A. by means of an automatic scan. Radiation transmitted by the cell was recorded using an RCA 93;i photocell. Corrections mere applied for variation in response of (1) Holder of a University of British Columbia Fellowship, 19621963, and a Kationnl Research Council Bursary, 1963-1964. (2) (a) 1'. B. Ayscough and E. W.R. Stencie, Proc. Roy. SOC.(London). A234, 476 (1956); (b) G . Gincometti. ti. Okabe, and E. W. I t . Stencie, ibid., A250, 287 (1959). (3) H. Oknhe nnd E. Vir, R. Steacie, Can. J . Chem., 36, 137 (1958). (4) (a) R. E. H u n t and W. A, Noyes, Jr., J . A m . Chem. Soc., 7 0 , 46 7 (1948); (b) P. Ausloos and E. Murad, J . Phys. Chern., 65, 1519 (1961).
( 5 ) D. J. LeRoy, Can. J . Res., B28, 492 (1950).
FLUORESCEKCE AND PHOSPHORESCENCE OF HEXAFLUOROACETONE VAPOR
both photocell and photomultiplier with wave length. Allowance was also made for change in the dispersion of the D285 spectrometer over the wave lengths recorded. The manufacturers' specifications mere used in considering these corrections.
2983
i
t
t
Results ( a ) Hexajluoroacetone. A typical emission trace as recorded is reproiduced in Fig. 1. The band a t 25' extends from below 3800 8. to a t least 7000 8. with a broad maximum between 4500 and 4900 8. Addition of a few nini. of oxygen decreases the intensity considerably and shifts the niaxiiiium to shorter wave lengths. KOfurther changes occur on introducing inore oxygen; it is probable that much less oxygen than was actually used completes the quenching. The foriii of the spectrum was essentially constant with ketone pressure between 30 and 150 nim. It would thus appear that even a t 25' a considerable part of the eniissilon is phosphorescence. The uinquenched fluorescerice band is almost identical with that previously r e p ~ r t e d but , ~ in that work there was apparently some impurity present which almost connpletely quenched the phosphorescence.6 In Fig. 2, where the corrections have been made for instrumental sensitivity, the two contributions to the total eiiiission are shown separately. The effect (of oxygen is even more marked at -78'. The phosphorescence has a niaxiiiiuni between 4900 and 5000 8. at both temperatures, and a t 25' there is a smaller peak near 4600 8. The fluorescence extends to 6000 8.,with a iiiaxiniuiii intensity close to 4300 8. a t each temperature. Table I shows the relative iiiagriitudes of the integrated intensities, I , found by measuring the area under appropriate corrected curves such as those in Fig. 2. Table I : Integrated Emission Intensities from HFA Vapor, and the Effect of Biacetyl and Oxygen; 3130-A. Radiation Was Used for Excitation, Except in Run 4B (4358 A,) Run
System
la b 2a b 3a
HFA (110 mm.) HFA 0 2 (1.9 mm.) HFA (30 mm.) HFA 02 (0.9 mm.) HFA (30 mm. ) HFA (30 mm.) HFA (33 mm.); biacetyl (14.7 mm.) HFA (33 mm.); biacetyl(14.7 mm.) HFA (33 mm. ) HFA biacetyl ( 2 mm.)
b 4a
b 5a b
+ +
+
T ,O C .
25
I
I(a)/I(b>
3 77
-78
10 95
-78 25 25
15 38
25
0 50
0 24
I
3800
4000
4200
4400 4600 4800 5000
WAVELENGTH
5500
6000
7000
(81
Figure 1. (a) Emission trace from 110 mm. of HFA a t 25'. ( b ) The eff:ct on mixing 1.9 mm. of 0 2 with ketone; 3130 A. was used for excitation.
2-
I
WAVELENGTH
(4)
Figure 2. ( a ) Phosphorescence and (b) fluorescence from 1, 110 mm. of HFA a t 25", and 2, 30 mm. of HFA a t -78". Relative intensities a t the different temperatures are not comparable.
( 6 ) HexajluoToacetone-Biacetyl M i x t 7 ~ r e ~ It . wts found that irradiation of HFA a t 25' with 3130 A. radiation in the presence of a m a l l concentration of biacetyl sensitized the strong characteristic green einission of biacetyl. In a mixture of 70 nini. of HFA with 1.3 nini. of biacetyl, the einission between 7000 and 4500 8. due to biacetyl wasJollowed by a much weaker "tail" extending to 3800 A. which corresponds closely to the fluorescence band of HFA. Neither fluorescenceonor phosphorescence from biacetyl extend below 4400 d . I (6) The emission observed here is not t h a t of hexnfluorobincetyl, present either as a trace impurity or as a product of the reaction. T h e emission spectrum of hexafluorobiacetyl has its principal maxima at 5400 and 5800 4. (7) H. Okabe and W. A. Noyes, Jr., J . Am. Chem. S o c , 7 9 , 801 (1957).
Volume 68, Number 10 October, 1564
2984
PETERG. BOWERSAND GERALDB. PORTER
I n run 4 (Table I), the concentrations of each coinponent were such that the iiiixture absorbed radiation to the saiw extent a t both 3130 8. (due to HFA) and 4358 8. (due to biacetyl). Hence, the direct and sensitized biacetyl einission intensities (4800-7000 A.) could be compared. A comparison was also made (run 5) of the intensity of the total HFA emission with that of the biacetyl eiiiission sensitized by the same sample.
Discussion We rewrite the sequence of primary events proposed
increases with decreasing temperature to a much greater extent than that of fluorescence
+ hv --+ ‘A*
IA*
A
(1)
dissociation
+ lA*
----)
lA0
(2)
+A
(3)
--+3Ao ‘A0
+A
‘A0 --f
3A0 SA0
---+A
(4)
+ hvf
(5)
A*
(6)
+ hv,
(7)
-+dissociation
(8)
3AO -+ A*
(9)
lA0 arid 3An represent excited singlet and triplet state niolecules, respectively, with no vibrational energy in cxccss of that required for thermal equilibration. The superscript * denotes vibrationally excited species. I t mill be convenient to associate quantum yields, +, with sonic of these processes. Thus, b2is the quanturn yield for dissociation froin ‘A*, +1 is the yield of formation of triplet molecules, and so on. In this riomcnclaturc, even if reaction 8, say, were perfectly efficient in the sense that all triplet molecules deconiposed, & need not necessarily be unity. According to this scheme, the ratio of phosphorescence to fluorescence is
67/65
== k h / ( k 7
+ + IC8
k 9 ) h
indcperidcrit of concentration. At the two tempcraturfs studied, this ratio (Table I) is (+7/&,)250= 2.77; (+7/45)--780 = 9.96 and, after allowing for a small change in absorption coefficient with temperature
(47
+
+5)--780/(+7
+
65)260
=
16.7
These figures show that the phosphorescence yield The Journal of Physical Chemistru
20.7;
(~6)-780/(#%)%0
=
5.75
This effect would be a t least partly due to the fact that triplet dissociation, which competes with phosphorescence a t 25O, is coiiipletelg inhibited a t the lower teinperaturc8 There are two paths by which transfer of electronic energy from HFA to biacetyl could occur, which result in sensitized emission froiii biacetyl
by Giaconietti, Oltabe, and Steacie,2b and include a phosphorescence step
A
=
(#7)--78O/(+7)26O
3Aa
+ B +3B0 + A 3Bo +B + h v ,
(10) (11)
and lA0
+B
---f
lB0
+A
(12)
1BO --+3Bo
(13)
+ hv,
--+ B
(11)
Coiiipleinentary studies of the dissociation of HFA under these conditions8 suggest that reaction 12 is much the less efficient, but that even a few tenths of a iiini. of biacetyl completely deactivates all triplet HPA molecules. The fact that the fluorescence band of HFA can still be observed in the presence of biacetyl supports this view. For purposes of the approxiniate calculations to follow, reaction 12 will